Our Privacy Statement & Cookie Policy

By continuing to browse our site you agree to our use of cookies, revised Privacy Policy and Terms of Use. You can change your cookie settings through your browser.

I agree

China flags security risks in OpenClaw open-source AI agent

CGTN

A graphic depicting the open-source AI agent OpenClaw. /VCG
A graphic depicting the open-source AI agent OpenClaw. /VCG

A graphic depicting the open-source AI agent OpenClaw. /VCG

China's information technology regulator on Thursday flagged security risks linked to the open-source AI agent OpenClaw.

China's National Vulnerability Database (NVDB), operated by the Ministry of Industry and Information Technology, said OpenClaw instances face elevated risks under default or improper configurations, making them vulnerable to cyberattacks and data leaks.

OpenClaw, formerly known as Clawdbot and Moltbot, is an open-source AI agent that integrates multi-channel communications with large language models to build customized assistants with persistent memory and proactive execution, and supports private, local deployment.

The agent has seen a viral rise since it was first introduced in November, receiving more than 100,000 stars on code repository GitHub and drawing in 2 million visitors in a single week, according to a blog post by its creator, Peter Steinberger.

It has also been gaining popularity among Chinese technology enthusiasts, with cloud service providers rushing to offer hosting solutions for the rapidly growing platform.

China's largest cloud service providers, including Alibaba, Tencent, and Baidu, have launched services that allow users to rent servers to run OpenClaw remotely, rather than on personal devices, according to the companies' OpenClaw deployment pages.

NVDB said unclear trust boundaries during deployment, combined with continuous operation, autonomous decision-making, and access to system and external resources, could expose instances to prompt-induced misuse, configuration flaws, or hostile takeovers in the absence of effective controls.

It also warned that such scenarios could lead to unauthorized actions, data leakage and system compromise if access controls, auditing and security hardening are insufficient.

The notice urged organizations and users to review public network exposure, permission settings and credential management, close unnecessary public access, and strengthen identity authentication, access control, data encryption and security auditing.

OpenClaw gained attention this week after a new social network, Moltbook, advertised that it is exclusively for OpenClaw bots. Cybersecurity firm Wiz said on Monday that the network had a major flaw that exposed private data on thousands of people.

(With input from Reuters)

Search Trends